| Power Amplifier(PA)plays an important role in wireless communication systems.However,when the PA is working in the saturation region,non-linear and memory effects often occur,which not only affects the work efficiency but also easily causes spectrum leakage.In order to improve the utilization of power resources,Digital PreDistortion(DPD)technology has become one of the mainstream methods of power amplifier linearization technology due to its low cost and good performance.Facing the call for large data transmission,the demand for multi-band and high-bandwidth signal are increasing.Compared with single-band transmitters,the distortion caused by concurrent multi-band transmitters is more complicated,and the impact on adjacent channels is also increased.Currently,there are many concurrent multi-band predistortion methods for linearizing multi-band power amplifiers.However,as the number of frequency bands continues to increase,the feedback path structure is also more complex.At the same time,the analog-to-digital converter(ADC)with a high sampling rate significantly increases the system cost.In order to reduce the system cost,simplify the structure of the digital predistortion feedback loop,and effectively reduce the sampling rate of the ADC,this paper proposes a novel multi-band digital predistortion method based on time-interleaved sampling based on a single feedback loop architecture.In the feedback loop,two time-synchronized ADCs are used to construct a co-prime undersampling loop to obtain time-interleaved sampling signals.According to the sparsity characteristics of concurrent multi-band signals in the frequency domain,and based on the theory of compressed sampling and signal reconstruction,the relationship between multi-band signals before and after compression is deeply explored.Then the multiband signals are reconstructed for digital predistortion.The experimental results show that compared with the traditional concurrent multi-band DPD method,the proposed method can still ensure the required linearization effect under the conditions of reducing the sampling rate and simplifying the feedback loop.In the meantime,the method also has good ability in terms of complexity and anti-noise performance. |